Saturday 15 March 2025
A new approach has been developed to identify individuals with mild cognitive impairment (MCI), a stage of brain degeneration that can precede Alzheimer’s disease. The method uses a combination of machine learning and functional magnetic resonance imaging (fMRI) to analyze the brain’s dynamic functional connectivity, which is the way different brain regions communicate with each other over time.
Current methods for diagnosing MCI rely on static measures of brain function, such as cognitive tests and structural MRI scans. However, these approaches may not accurately capture the complex changes that occur in the brain as it declines. The new method, developed by researchers at the University of Texas at Arlington, uses a technique called dynamic functional connectivity (dFC) to analyze the brain’s activity over time.
The researchers used fMRI data from 345 participants, including those with MCI and healthy controls, to develop their approach. They created a network of brain regions that were connected by functional fibers, which represented the strength of communication between each pair of regions. The network was then analyzed using a machine learning algorithm to identify patterns of connectivity that distinguished between individuals with MCI and those without.
The results showed that the new method was able to accurately identify individuals with MCI, even when they did not show obvious signs of cognitive decline. The approach also outperformed traditional methods in identifying individuals who would eventually develop Alzheimer’s disease.
The researchers believe that their method could be used as a diagnostic tool for MCI, potentially allowing for earlier intervention and treatment. They also suggest that the approach could be applied to other brain disorders, such as Parkinson’s disease and depression.
One of the key advantages of the new method is its ability to capture complex patterns of brain activity over time. This is because fMRI data can provide a detailed picture of how different brain regions communicate with each other, even when there are no obvious changes in brain structure or function.
The researchers used a combination of machine learning and graph theory to analyze the fMRI data and identify patterns of connectivity that distinguished between individuals with MCI and those without. They also developed a novel algorithm that was able to adapt to individual differences in brain activity, allowing for more accurate diagnosis.
The study provides new insights into the complex changes that occur in the brain as it declines, and could potentially lead to the development of new treatments for MCI and Alzheimer’s disease.
Cite this article: “New Method Identifies Mild Cognitive Impairment Using Machine Learning and fMRI”, The Science Archive, 2025.
Machine Learning, Functional Magnetic Resonance Imaging, Mild Cognitive Impairment, Dynamic Functional Connectivity, Brain Degeneration, Alzheimer’S Disease, Diagnostic Tool, Fmri Data, Graph Theory, Neural Networks.
